Proximal Renal Tubular Dysfunction Related to Antiretroviral Therapy Among HIV-Infected Patients in an HIV Clinic in Mexico

Abstract

Proximal renal tubular dysfunction (PRTD) of varying severity has been associated with antiretroviral toxicity, especially related to the use of tenofovir (TDF). The aim of this study was to investigate whether HIV-infected patients who use a tenofovir-based regimen are at increased risk of tubular dysfunction. We conducted an observational, comparative, longitudinal, prospective study. Estimated glomerular filtration rate (eGFR) and markers of tubular damage to assess tubular dysfunction (fractional excretion of phosphate and uric acid, glycosuria, and proteinuria) were measured at baseline and at weeks 12 and 24. Of 111 participants, PRTD was found in 6.3% at week 12 and 9% at week 24, with no statistically significant difference between those on an abacavir (ABC)-containing regimen or a TDF-containing regimen. We also found an increase in triglycerides associated with the ABC-containing regimen compared with the TDF group. The use of an ABC- or TDF-containing regimen was independently associated with tubular dysfunction, but we found no significant differences between these groups, except when TDF was combined with a protease inhibitor. A better and more complete assessment of renal function is needed, because the presence of tubular dysfunction and proteinuria without impairment of eGFR may affect the renal safety of HIV-infected patients.

Introduction

The use of highly active antiretroviral therapy (HAART) has improved the prognosis of patients infected with human immunodeficiency virus (HIV).¹ In this new era of HIV care management, non-acquired immunodeficiency syndrome (AIDS)-related complications of HIV infection have become increasingly important.² Among these, kidney disease is relatively common, with a prevalence of chronic kidney disease between 5% and 15% in this population,^3-–5 and proteinuria being detected in up to 30% of patients.^6,7 Many of these renal problems have been associated with toxicity of HAART, especially relating to tenofovir (TDF)-based regimens in combination with protease inhibitors (PIs).⁸ Renal toxicity may lead to features of proximal renal tubular dysfunction (PRTD), acute kidney injury, isolated proteinuria, or chronic kidney disease.⁹ TDF is the drug most often associated with tubular alterations, with a reported incidence of Fanconi syndrome of 0.3–2%,¹⁰ and a prevalence of PRTD (defined as two or more tubular alterations) of 6–15%,^2,11,12 and up to 49% with combined TDF/PI treatment. Hypophosphatemia has been reported in up to 30% of HIV-infected patients, although multiple factors are often implicated, and decreased phosphate tubular reabsorption has been reported in 4–50%.¹¹

Normoglycemic glycosuria has been found in 2% of TDF-treated patients with a normal estimated glomerular filtration rate (eGFR).¹³ The clinical significance of isolated tubular abnormalities in the short- and long-term remains unclear, and it is unknown whether these patients are at increased risk of Fanconi syndrome or progressive alteration of GFR. These tubular abnormalities may be missed until they affect glomerular function, so earlier and more accurate biomarkers of PRTD are needed for monitoring renal function in HIV-infected patients under HAART, especially those treated with TDF-based regimens alone and in combination with PIs. However, in our country, we rarely use the newer biomarkers that have been developed for this purpose. There are no recent studies of tubular renal markers in HIV-infected patients in the Mexican population. The aim of our study was to investigate whether Mexican HIV-infected patients who use a TDF-based treatment regimen are at an increased risk of tubular dysfunction.

Methods

Study design

After obtaining approval from the ethics and investigation committee in “La Raza” Medical Center number 35021, we conducted an observational, comparative, prospective, longitudinal, open-labeled study that was carried out between August 2012 and December 2013 at the National Medical Center “La Raza” in Mexico City.

Patients

The study sample included treatment-naïve HIV-infected patients, aged between 18 and 50 years, who had indications to start treatment (CD4 count below 350 cells/μL or above this level with co-morbidities), with eGFR above 90 mL/min per 1.73 m², and who attended the HIV clinic of our National Health Service. We excluded patients with chronic diseases including diabetes mellitus, hypertension, previous renal disease, obstructive uropathy, hyperuricemia, cardiovascular disease, neoplasms, hepatitis C co-infection, prior proteinuria (urinary protein creatinine ratio UP/C over 150 mg/g), impaired markers of tubular dysfunction prior to the start of HAART, and patients who had bone or calcium–phosphorus metabolism diseases. Other exclusion criteria were a decline in eGFR of 25% or more during the period following a switch in antiretroviral regimen, suspicion of nephrotoxicity from another cause, and death. Patients were also excluded if they withdrew their informed consent or if they lost National Health Service benefits during the time of the study. After obtaining informed consent, each participant was interviewed, and information including age, current use of tobacco, alcohol, trimethoprim or other drugs, chronic diseases, co-infections, and nephrourologic disorders was recorded on the designated record forms.

Measurements

The enrolled patients' medical records were reviewed for data including the duration of HIV infection, current CD4 cell count, and current plasma HIV-1 RNA level. Baseline serum levels of glucose, urea, creatinine, albumin, cholesterol, triglycerides, uric acid, phosphorus, and calcium were assessed; urine spot quantitative measures of creatinine, uric acid, phosphate, and protein were made to determine the fractional excretion rates (FE) and the protein/creatinine ratio (UPC), respectively. Urinalysis to assess pH, urine specific gravity, glycosuria, and urine sediment was also performed. The eGFR was calculated using the Modification of Diet and in Renal Disease-6 (MDRD-6) to check that this was above 90 mL/min. If any alterations in the previously recorded levels were found, the inclusion of the patient was complete. Core therapy was started with either TDF/emtricitabine (TDF-based) or abacavir/lamivudine (ABC-based) regimens according to the physician's decision; HIV-1 RNA viral load was one important factor in making this decision. Serum and urinary markers and FE and UP/C were assessed at weeks 12 and 24. PRTD was considered to be present with the presence of two or more impaired markers of tubular dysfunction: normoglycemic glycosuria, fractional excretion of phosphate (FEPh) >20% or >10% if serum phosphate <2.5 mg/dL, FE of uric acid (FEUA) >10%, or UP/C >150 mg/g. A significant impairment in eGFR was defined as a 25% decline from the baseline level.

Statistical analysis

Statistical analysis was performed using the Statistical Package for Social Sciences (SPSS) software, version 21. Descriptive statistics were used to report demographics and were presented in contingency tables for categorical variables and continuous variables. For non-normally distributed continuous variables, median and 25–75 interquartile range were reported; for normally distributed variables, mean and standard deviation were used. Univariate analysis to compare variables in patients with and without tubular dysfunction was performed using a chi-squared test, and for continuous variables a t-test or Mann–Whitney U test was used depending on the observed distribution. Statistical significance for all analyses was defined as a p value<0.05.

Results

A total of 111 HIV-infected treatment-naïve Mexican patients were enrolled. The baseline descriptive characteristics of the participants are shown in Table 1. The mean age was 31.05±8 years, and 91.9% (102) of the patients were men. Seventy (63.1%) were given a TDF-based regimen and 41 (36.9%) an ABC-based regimen; in each of these groups there were 11 patients who also received a protease inhibitor (PI), lopinavir/ritonavir. The mean eGFR was 118.4±20 mL/min; however, in 23.4% of the study group the eGFR was at levels indicating hyperfiltration (eGFR above 130 mL/min in men and 120 mL/min in women). No baseline differences were found between the TDF and ABC groups except for plasma HIV-1 RNA levels (p<0.001).

Table 1.

Baseline Characteristics of Participants

	TDF group n=70	ABC group n=41	p Value
Age in years (mean±SD)	31.2±8.4	30.6±7.3	0.71
Male no. (%)	66 (94.2)	36 (87.8)	0.28
BMI (Mean±SD)	23.6±2.5	23.4±2.6	0.65
Glucose, mg/dL (mean±SD)	88.8±7	89.8±8.9	0.52
Urea, mg/dL (mean±SD)	26.6±8.1	26.15±7.9	0.76
Median serum creatinine, mg/dL (IQR 25–75)	0.86 (0.8–0.92)	0.81 (0.77–0.9)	0.07
Median serum uric acid, mg/dL (IQR 25–75)	5.67 (5.07–6.4)	5.7 (4.51–6.14)	0.45
Cholesterol, mg/dL (mean±SD)	152.6±43	139.4±38.3	0.10
Median triglycerides, mg/dL (IQR 25–75)	141.5 (108.75–177)	137 (111–196)	0.91
Median albumin, mg/dL (IQR 25–75)	4.3 (3.97–4.4)	4.3 (4.0–4.5)	0.77
Median serum phosphorus, mg/dL (IQR 25–75)	3.6 (3.2–3.9)	3.5 (3.15–3.85)	0.4
Median serum calcium, mg/dL (IQR 25–75)	9.25 (9.0–9.77)	9.2 (8.95–9.6)	0.19
Median urine creatinine, mg/dL (IQR 25–75)	75.27 (59.92–100.39)	84.3 (65.89–116.83)	0.07
Median urine phosphate, mg/dL (IQR 25–75)	22.99 (18.85–33.1)	27.5 (18.07–41.21)	0.27
Median urine uric acid, mg/dL (IQR 25–75)	18.3 (12.82–23.77)	18.7 (12.75–23.3)	0.90
Median HIV-1 RNA, copies/mL (IQR 25–75)	124,000 (56,589–231,000)	42,100 (27,072–76,110)	<0.001^a
Median CD4+count, cells/mm³ (IQR 25–75)	227.75 (165.55–323.05)	287 (149.25–329)	0.4
No. (%) taking protease inhibitor	11 (15.7)	11 (26.8)	0.2
Median FEPh, % (IQR 25–75)	7.24 (5.89–10.75)	7.45 (5.81–9.85)	0.6
Median FEUA, % (IQR 25–75)	3.9 (2.63–5.07)	3.24 (2.25–4.4)	0.22
MDRD eGFR, mL/min (mean±SD)	118.99±23.1	118.85±21.72	0.90

ABC, abacavir; eGFR, estimated glomerular filtration rate; FEPh, fractional excretion of phosphate; FEUA, fractional excretion of uric acid; IQR, interquartile range; SD, standard deviation; TDF, tenofovir; ^a p value<0.05.

At week 12, no patients showed normoglycemic glycosuria; hypophosphatemia (serum phosphorus<2.5 mg/dL) was observed in 5 (4.5%) patients. Fourteen patients (12.6%) had impaired FEPh, 11 (9.9%) impaired FEUA, and 15 (13.5%) had significant proteinuria. We found a significant association between smoking and impaired FEPh (p=0.011), and baseline hypercholesterolemia and altered FEUA (p=0.014). No significant associations were found for treatment group regimen, initial CD4 count, HIV-1 RNA level, hepatitis B virus co-infection, or sex. PRTD (2 or more impaired markers) was observed in 7 (6.3%) of 111 patients; 4 of these were being treated with a TDF-based regimen (p=0.708). Tubular dysfunction was significantly associated with baseline hypercholesterolemia (p=0.04). When we compared parameters of serum and urinary markers at week 12 between TDF and ABC groups, we found significant differences between the two groups in serum creatinine (p=0.043), triglycerides (p=0.039), urine creatinine (p=0.029), and FEUA (p=0.024) (Table 2). At week 24, 9 patients (8.1%) had hypophosphatemia (6 of whom used the TDF-based regimen), 22 (19.8%) had an impaired FEPh, 16 (14.4%) an altered FEUA, and 23 (20.7%) presented with significant proteinuria. A significant association was found between proteinuria and initial HIV-1 RNA level (p=0.026). Impaired FEUA was associated with baseline hypercholesterolemia (p=0.007), and PI treatment (p=0.032). PRTD was found in 10 (9%) of the patients, but no significant associations of PRTD with other variables were found (Table 3). A significant decline in eGFR was found in 4 (3.6%) patients, 3 of whom were being treated with the TDF-based regimen (p=1.0). Two of these 3 patients were also treated with a PI, and this combination treatment was significantly associated with a decline in eGFR (p=0.048). Significant proteinuria at week 24 was also significantly associated with a decline in eGFR (p=0.028).

Table 2.

Results at Week 12

	TDF group n=70	ABC group n=41	p Value
Glucose, mg/dL (mean±SD)	92±9	91.73±9.5	0.883
Urea, mg/dL (mean±SD)	25.77±7.28	25.98±7.07	0.883
Median serum creatinine, mg/dL (IQR 25–75)	0.87 (0.82–0.94)	0.82 (0.77–0.89)	0.010^a
Median serum uric acid, mg/dL (IQR 25–75)	5.63±1.18	5.63±1.49	1.0
Cholesterol, mg/dL (mean±SD)	159.5 (33–182.5)	145 (123–200.5)	0.443
Median triglycerides, mg/dL (IQR 25–75)	154.5(123.5–200.25)	178 (120–325.5)	0.039^a
Median albumin, mg/dL (IQR 25–75)	4.35 (4.1–4.5)	4.30 (4.15–4.6)	0.963
Median serum phosphorus, mg/dL (IQR 25–75)	3.4 (.0–3.8)	3.5 (3.1–3.8)	0.799
Median serum calcium, mg/dL (IQR 25–75)	9.5 (9.1–9.7)	9.3 (9.0–9.5)	0.024^a
Median urine pH median (IQR 25–75)	6.0 (6.0–6.5)	6.0 (6.0–6.0)	0.317
Median specific gravity (IQR 25–75)	1.017 (1.010–1.020)	1.019 (1.018–1.021)	0.020 ^a
Median urine creatinine, mg/dL (IQR 25–75)	78.8 (65.58–102.17)	91. 68 (62.25–129.3)	0.135
Median urine phosphate, mg/dL (IQR 25–75)	37.55 (25–54.4)	43.5 (21.68–65.05)	0.753
Median urine uric acid, mg/dL (IQR 25–75)	26.35 (18.15–35.5)	21.8 (16.9–34.85)	0.318
Median FEPh, % (IQR 25–75)	12.05 (9.38–16.43)	9.76 (6.93–14.93)	0.072
Median FEUA, % (IQR 25–75)	5.5 (3.65–7.92)	4.24 (2.92–6.55)	0.040
MDRD eGFR, mL/min (mean±SD)	118.59±23.28	119.67±23.1	0.814

Table 3.

Results at Week 24

	TDF group n=70	ABC group n=41	p Value
Glucose, mg/dL (mean±SD)	91 (86.75–99)	90 (85–96.5)	0.910
Urea, mg/dL (mean±SD)	26.15±6.69	25.83±8.04	0.823
Median serum creatinine, mg/dL (IQR 25–75)	0.86 (0.80–0.90)	0.84 (0.78–0.92)	0.434
Median serum uric acid, mg/dL (IQR 25–75)	5.53±1.10	5.5±1.2	0.898
Cholesterol, mg/dL (mean±SD)	173.23±39.47	179.49±46.8	0.456
Median triglycerides, mg/dL (IQR 25–75)	163 (123–233.25)	206 (165–60)	0.013^a
Median albumin, mg/dL (IQR 25–75)	4.35 (4.2–4.52)	4.4 (.43–4.5)	0.657
Median serum phosphorus, mg/dL (IQR 25–75)	3.2 (2.9–3.6)	3.4 (3.0–3.9)	0.075
Median serum calcium, mg/dL (IQR 25–75)	9.5 (9.07–9.7)	9.4 (9.05–9.6)	0.338
Median urine pH median (IQR 25–75)	6.0 (6.0–6.5)	6.0 (6.0–6.5)	0.652
Median specific gravity (IQR 25–75)	1.016 (1.010–1.022)	1.015 (1.010–1.022)	0.924
Median urine creatinine, mg/dL (IQR 25–75)	75.15 (61.19–97)	79.25 (64.14–101.44)	0.327
Median urine phosphate, mg/dL (IQR 25–75)	40.15 (27.62)	41.85 (30.4–57.75)	0.867
Median urine uric acid, mg/dL (IQR 25–75)	26.5 (17.75–35.05)	29.80 (19.1–39.3)	0.292
Median FEPh, % (IQR 25–75)	14.8(11.32–19.19)	11.69 (9.54–19.06)	0.271
Median FEUA% (IQR 25–75)	5.11 (3.68–7.53)	5.54 (3.37–7.56)	0.845
MDRD eGFR, mL/min (mean±SD)	119.55±20.4	117.65±24.02	0.658

Discussion

PTRD (defined by the presence of at least two of glycosuria, proteinuria, impaired FEPh, or impaired FEUA), was found in 6.3% and 9% of patients at weeks 12 and 24, respectively, with no significant difference between the groups treated with the ABC-containing regimen or the TDF-containing regimen.

We also found an increase in triglycerides in the group treated with the ABC-containing regimen compared with the TDF-treated group.

The incidence of PTRD is comparable with that reported in other trials in HIV-infected populations in other countries, which ranges from 2% to 15%.^2,11,12 PRTD has been reported in up to 22% of TDF-treated patients in different cohorts⁸ with a longer follow-up of patients. However, in our study we found an incidence of 10% in this group of Mexican patients at 24 weeks. The lack of a significant difference between treatment groups might suggest that a larger sample size or longer follow-up of patients is needed, because in some cohorts, significant differences between treatment groups have been reported at 6 months and beyond.¹⁴ Only proteinuria was significantly associated with the presence of PRTD at week 24. This can be explained as tubular proteinuria, as no other significant associations were demonstrated in this population.

Renal phosphate wasting is possibly the most sensitive marker of tubular dysfunction, and is one of the features of Fanconi syndrome and PRTD. Hypophosphatemia was observed in 4.5% and 8.1% of all patients at weeks 12 and 24, respectively. We did not observe significant differences in median serum phosphorus levels between our study groups, but there was a nonsignificant increase in the proportion of patients with hypophosphatemia in the TDF group, which should be explored further in larger studies. We observed a significant increase in FEPh from baseline (p<0.001) at 12 weeks that persisted to the end of follow-up, without significant differences between study groups. Impaired FEPh (defined as >20% or >10% if serum phosphate <2.5 mg/dL) was found in 14 patients (12.6%) at week 12, and 22 patients (19.8%) at week 24, with a nonsignificantly higher proportion in the TDF group at the end of the study. In our study, impaired renal phosphate wasting was present in 12.8% of the TDF group vs. 4.8% of patients not treated with TDF, in contrast with studies that report a prevalence of 42–50% in TDF-treated patients.¹¹ Smoking (at week 12) and proteinuria (at week 24) were associated significantly with impaired FEPh, which can be explained by vascular or tubulointerstitial damage in these patients. No other parameters were associated with this alteration.

The kidney plays a major role in uric acid homeostasis, being responsible for more than 70% of its excretion as urate, which predominantly occurs in the proximal tubule.¹⁵ The normal FEUA is about 10%, and higher levels can indicate renal wasting, which is why we considered that it could be a marker of tubular dysfunction. Impaired FEUA was found in 12.8% and 14.2% of patients in the TDF group compared with 4.8% and 14.6% in the non-TDF group at weeks 12 and 24, respectively, but this difference was not statistically significant. However, FEUA levels differed significantly between treatment groups at week 12, although they reached similar levels at week 24. This can be explained by the modification of the purines in the patients' diets when altered levels were detected. Rates of proteinuria were high in patients with HIV in previous studies,¹⁶ but no significant differences were found between different treatment groups.

A limitation of our study was that we did not determine the levels of urine β2-microglobulin or retinol binding protein (RBP), which are more specific measures to assess tubular proteinuria; instead, we measured total proteinuria. A significant association was found between baseline HIV-1 RNA levels and proteinuria (p=0.026) at week 24 in bivariate analysis, which is consistent with previous studies.¹⁷ The importance of detection of proteinuria or microalbuminuria in HIV-infected patients is that it has been associated with an increased risk of mortality even after adjustment for markers of HIV disease severity,¹⁸ so it should be assessed further in our study population.

Identification of the causes of a significant fall in eGFR has been the objective of many previous trials. Cumulative TDF use and ritonavir-boosted PI use have been associated with this feature.¹⁹ In our study, baseline eGFR indicated that 23.4% of patients showed hyperfiltration, with a mean of 147.99±12.15 mL/min. Our data report a mean fall in eGFR of −1.24 mL/min in the TDF group at week 24, in contrast to an increase of 1.13 mL/min in the non-TDF group (p=0.43). Consistent with previous studies, we found a significant association between the presence of proteinuria, use of a TDF+PI regimen, and a significant fall in eGFR. This could have several explanations. PI coadministration with TDF is known to increase TDF serum levels in 20–30% of patients, reducing its clearance by interfering with renal excretion transporters. Another suggested mechanism is that this combination might result in a greater TDF-induced inhibition of creatinine secretion in the proximal tubules rather than a greater decline in eGFR; this could also explain the significantly higher levels of serum creatinine that we found at week 12. For this reason, other methods should be used to assess a significant fall in eGFR, such as radionuclide determination of GFR, and further studies need to be performed in this setting.

In conclusion, among HIV-positive persons, use of ABC- or TDF-containing regimens was independently associated with tubular dysfunction. However, we did not find significant differences between these two treatment groups except when TDF was combined with a PI. A better and more complete assessment of renal function in HIV patients is needed, because tubular dysfunction and proteinuria without impairment of eGFR may affect the renal safety of HIV-infected patients.

Footnotes

Acknowledgments

The authors wish to acknowledge all patients who kindly gave consent to participate in this study. The authors also wish to thank the study staff in the Nephrology and Infectious Diseases Departments and the study participants and their families and caregivers for participation in the trial.

Author Disclosure Statement

All authors declare no conflict of interest.

References

Palella

Jr , Delaney

, Moorman

, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N Engl J Med, 1998; 338:853–860.

Dauchy

, Lawson-Ayayi

, de la Faille

, et al. Increased risk of abnormal proximal renal tubular function with HIV infection and antiretroviral therapy. Kidney Int, 2011; 80:302–309.

Szczech

, Hoover

, Feldman

, et al. Association between renal disease and outcomes among HIV-infected women receiving or not receiving antiretroviral therapy. Clin Infect Dis, 2004; 39:1199–1206.

Wyatt

, Winston

, Malvestutto

, et al. Chronic kidney disease in HIV infection: An urban epidemic. AIDS, 2007; 21:2101–2103.

Mocroft

, Kirk

, Gatell

, et al. Chronic renal failure among HIV-1-infected patients. AIDS, 2007; 21; 1119–1127.

Gupta

, Mamlin

, Johnson

, et al. Prevalence of proteinuria and the development of chronic kidney disease in HIV-infected patients. Clin Nephrol, 2004; 61:1–6.

Gardner

, Holmberg

, Williamson

, et al. Development of proteinuria of elevated serum creatinine and mortality in HIV-infected women. J Acquir Immune Defic Syndr, 2003; 32:203–209.

Del Palacio

, Romero

, Casado

. Proximal tubular renal dysfunction or damage in HIV-infected patients. AIDS Rev, 2012; 14:179–187.

Maggi

, Bartolozzi

, Bonfanti

, et al. Renal complications in HIV disease: Between present and future. AIDS Rev, 2012; 14:37–53.

10.

Jones

, Stebbing

, Nelson

, et al. Renal dysfunction with tenofovir disoproxil fumarate-containing highly active antiretroviral therapy regimens is not observed more frequently: A cohort and case-control study. J Acquir Immune Defic Syndr, 2004; 37:1489–1495.

11.

Fux

, Opravil

, Cavassini

, et al. Tenofovir and PI use are associated with an increased prevalence of proximal renal tubular dysfunction in the Swiss HIV Cohort Study: Boosted IP+TDF associated with renal function. 16th CROI, 2009, Montreal, Canada (Abstract 743).

12.

Labarga

, Barreiro

, Martin-Carbonero

, et al. Kidney tubular abnormalities in the absence of impaired glomerular function in HIV patients treated with tenofovir. AIDS, 2009; 23:689–696.

13.

Herlitz

, Mohan

, Stokes

, et al. Tenofovir nephrotoxicity: Acute tubular necrosis with distinctive clinical, pathological, and mitochondrial abnormalities. Kidney Int, 2010; 78:1171–1177.

14.

Maggi

, Montinaro

, Bellacosa

, et al. Early markers of tubular dysfunction in antiretroviral-experienced HIV-infected patients treated with tenofovir versus abacavir. AIDS Patient Care STDS, 2012; 26:5–11.

15.

Lipkowitz

. Regulation of uric acid excretion by the kidney. Curr Rheumatol Rep, 2012; 14:179–188.

16.

Kabanda

, Vandercam

, Bernard

, et al. Low molecular weight proteinuria in human immunodeficiency virus-infected patients. Am J Kidney Dis, 1996; 27:803–808.

17.

Szczech

, Gange

, van der Host

, et al. Predictors of proteinuria and renal failure among women with HIV infection. Kidney Int, 2002; 61:195–202.

18.

Wyatt

, Hoover

, Shi

, et al. Microalbuminuria is associated with all-cause and AIDS mortality in women with HIV infection. J Acquir Immune Defic Syndr, 2010; 55:73–77.

19.

Ryom

, Mocroft

, Kirk

, et al. Association between antiretroviral exposure and renal impairment among HIV-positive persons with normal baseline renal function: The D:A:D study. J Infect Dis, 2013; 207:1359–1369.